Creating a Csv using Powershell - csv

Question

I am trying to take a filename such as: John_Doe_E_DOB_1/1/46_M(This is the gender)_ID_0000000_IMG_FileName_Date-of-File_1/1/15_Doc-page-1 And create a CSV file to open in Excel with column headers for: Last Name, First Name, MI, ID No, File Name, Date of File along with doc type. Here's my code so far:
Get-ChildItem -Path C:\Users\name\desktop\test -Recurse | ForEach-Object {$_ | add-member -name "Owner" -membertype noteproperty -value (get-acl $_.fullname).owner -passthru} | Sort-Object fullname | Select BaseName,Name,Owner | Export-Csv -Force -NoTypeInformation C:\Users\name\desktop\test\thing.csv
All this is doing is dropping that really long file name in at the top, and then adding the ext at the end in another column. Example:
John_Doe_E_DOB_1/1/46_M(This is the gender)_ID_0000000_IMG_FileName_Date-of-File_1/1/15_Doc-page-1 Would be in column 1 and
John_Doe_E_DOB_1/1/46_M(This is the gender)_ID_0000000_IMG_FileName_Date-of-File_1/1/15_Doc-page-1.txt <----- Would be the only difference in column 2
How can I split this up for over a million files, all different lengths, and sizes, and get it to break up into the categories listed above? All help would be greatly appreciated.

Answer 1

I would replace the Select stage of your pipeline with a call to a filter function like this:
filter GenObj {
$parts = $_.FullName.Split('_')
new-object pscustomobject -property #{
Owner = (get-acl $_.fullname).owner
FirstName = $parts[0]
LastName = $parts[1]
MiddleInitial = $parts[2]
# Fill in the rest
}
}
Get-ChildItem -Path C:\Users\name\desktop\test -Recurse |
Sort-Object fullname |
GenObj |
Export-Csv -Force -NoTypeInformation C:\Users\name\desktop\test\thing.csv
This will create a new custom object with all the properties on it that correspond to the parts of the filename you want to extract.
This string splitting approach may not work depending on how you handle names with no middle initial.
Also be aware that if you are processing a million files, the use of Sort-Object will cause every single FileInfo object (one for every file) to get buffered in memory so the sort can be performed. You may likely run out of memory and the command will fail. I would consider removing Sort-Object in this scenario.